Schizophrenia is a highly heritable neuropsychiatric disorder with significant public health costs. Understanding the contributions of genetic variability in the disease will help identify better predictors of prognosis and treatment response Current studies are using genome wide scan (GWS) approaches to identify the numerous genes which might play a role in schizophrenia-either in increased risk for the disorder overall, or through modulating the various clinical symptoms. Structural neuroimaging measures implicate gray matter loss in schizophrenia; subjects with schizophrenia tend to have larger ventricles and smaller grey matter volumes than do their healthy counterparts, and regional loss in the medial frontal, temporal and insular gyri have been identified by us and others. Identifying the genetic influences underlying these patterns of gray matter loss is the goal of this project. We propose a multivariate method for analyzing already existing GWS data and voxelwise measures of gray matter density. We will apply parallel ICA, with reference; this technique identifies patterns of spatial variation in the brain structure and patterns of genotypes which are linked. We begin with 3200 structural imaging and GWS samples from healthy controls and schizophrenics, from aggregated legacy data. We constrain the imaging and genetic analyses with reference vectors to incorporate a priori information. In Aim 1 we will develop initial a prioi spatial patterns, structural networks using source- based morphometry methods; in Aim 2 we will determine the heritability and quantitative trait loci for these networks in independent famil samples; in Aim 3 we plan to use the quantitative trait loci as a priori constraints on the genetic data, and the heritable structural networks as constraints on the imaging data on our parallel ICA analysis. Using these methods, we will determine the spatial patterns and genetic profiles that covary within our sample, and which show effects of diagnosis in schizophrenic and control data. We include a split-half analysis for replication and a follow-up high-density genotyping plan. This approach identifies structural networks within the brain allowing for variation in age, medication exposure, and other measures, and links them to genetic combinations. The final results will be the combinations of genotypic profiles which influence the patterns of structural brain loss in the disease. The methods developed will allow complex imaging and GWS data to be analyzed in combination, potentially applicable to many disorders.